TigerGraph launches Workbench for graph neural network ML/AI modeling

TigerGraph, creator of the Graph Analytics platform for data scientists, today unveiled its TigerGraph ML (Machine Learning) workbench during its Graph & AI Summit event, a new-generation toolkit that will significantly improve the accuracy of ML models to analysts. .

Workbench does this while using familiar tools, workflows and libraries in a single environment that plugs directly into existing data pipelines and ML infrastructure, ”said Victor Lee, VP of TigerGraph.

ML Workbench is a Jupiter-based Python development framework that enables data scientists to create deep-learning AI models using data connected directly from the business. Graph-capable MLs have been shown to have more accurate predictive power and take much less time than traditional ML approaches.

Traditional machine learning algorithms are based on the learning of the system through training sets to develop trained models. This pre-trained model is used to classify or identify test datasets; This can usually take days or weeks to finalize for a specific use case. Graph-based MLs can sometimes take minutes to create algorithmic models.

The value of ML is high, but so is the learning curve

“Graph ML has proven to accelerate and improve learning and performance, but the curve of learning to use APIs (application programming interfaces) and libraries to make it happen has proven to be very strong for many data scientists,” Lee said in a media advisory. Was. “We have therefore created the ML workbench to provide a new functional level between data scientists and graph machine-learning APIs and libraries to facilitate data storage and management, data preparation and ML training.

“In fact, we’ve seen early adopters gain a 10-50% increase in the accuracy of their ML models as a result of using ML workbenches and TigerGraph,” he said.

Lee told VentureBeat that TigerGraph’s whole thinking revolves around the definition of human identity, based on how you interact with others.

“The same thing is true with graphs in data modeling, and it is now expanding to neural networks.” Lee said. “Each node in the graph is interconnected like people. Graphs are excellent for asking pattern-matching algorithms. Workbench will help you use machine learning based on the information inside the graph, but the real power graph comes with neural networks, which are regular graphs on steroids.

“Our DGL (Deep Graph Library), for example, there is an extension of (Metana) Pyotorch geometry that supports graph neural networks,” he said. “This is a great feature, and it shows that we are going where the data scientists are; We are not trying to teach them something new. We’re using tools that they already know and are comfortable with, because we’re trying to reduce the learning curve. “

Best for cases of fraud, predictive use

The ML workbench enables organizations to determine improved insights into node-prediction applications, such as fraud and edge-prediction applications, including product recommendations, Lee said. The ML workbench enables AI / ML practitioners to explore graph-enhanced machine learning and graph neural networks (GNNs) as it is fully integrated with TigerGraph’s database for parallel graph data processing / manipulation, Lee said.

The ML workbench is designed to interact with popular deep learning frameworks such as Pytorch, Pytorch Geometric, DGL and Tenserflow, giving users the flexibility to choose the framework they are most familiar with. The ML workbench is also plug-and-play ready for Amazon Sagemaker, Microsoft Azure ML and Google Vertex AI, Lee said.

The ML workbench is designed to work with enterprise-level data. Users can train GNN because of the following built-in capabilities – even on very large graphs:

  • Of Tigergraph DB Distributed storage and large-scale parallel processing;
  • Graph based partition Training / validation / testing to generate graph data sets;
  • Graph based batching To improve performance and reduce HW requirements for GNN mini-batch training; And
  • Subgraph Sampling Leading edge to support GNN modeling techniques.

ML Workbench is compatible with TigerGraph 3.2 Forward, available as a fully powered cloud service and for on-premises use. Currently available as a preview, the ML workbench will normally be available in June 2022, Lee said.

TigerGaph competes with Neo4J, ArangoDB, MemGraph and a few others in the graph database space.

Selection of Winners of ‘Million Dollar Challenge’

At the Graph & AI Summit, TigerGraph unveiled the winners of the Graph for All Million Dollar Challenge – game-changing, graph-driven projects that address and address today’s largest global social, economic, health and analysis આપવામાં 1 million in cash. Is. Climate concerns.

The winning projects announced at the Graph + AI Summit this week were hand-picked by a global judging panel from more than 1,500 entries from more than 100 countries. Mental Health Hero has claimed a પ્રા 250,000 grand prize for creating an application to help provide greater access and personalization to mental health treatment.

Similar Posts

Leave a Reply

Your email address will not be published.